🛠️ delegate-team

Routes coding tasks through the right agentic workflow, governed by local memory, rules, and policy checks.

One CLI. One orchestrator. One optional memory + governance layer. Pick what you need, skip the rest.

⚡ Pick your lane

| If you want... | Lane | What gets installed | Time | |---|---|---|---| | Just the dt CLI to dispatch tasks to backends | Lane 1 — dt CLI | npm package → dt on PATH | 30 s | | /mavis-ship "X" slash command in Claude Code | Lane 2 — /mavis-ship | Lane 1 + orchestrator skill | 2 m | | The full local agent OS (memory, multi-agent, governance) | Lane 3 — full local agent OS | Lane 2 + MMAS + agent-kernel | 5–10 m |

Not sure? Most users start with Lane 1. You can always add Lane 2 / 3 later — every install is idempotent.

See docs/INSTALLATION.md for the full reference.

Lane 1 — dt CLI only

npm install -g delegate-team
dt --version          # → 2.7.1
dt run "<task>"       # dispatch with auto backend selection
dt doctor             # health check

What you get: one Node CLI. No skills, no hooks, no companion frameworks, no network calls at runtime.

Uninstall: npm uninstall -g delegate-team

Lane 2 — /mavis-ship in Claude Code

npm install -g delegate-team                     # Lane 1 first
git clone https://github.com/imMamdouhaboammar/delegate-team
cd delegate-team
./install.sh --orchestrator                      # install the skill
# Restart Claude Code so /mavis-ship is registered

Then in any Claude Code session:

/mavis-ship "Build a CSV → JSON CLI"
mavis-orchestrate "ship v0.2.0 to npm"           # routing decision only
dt route --explain "<task>"                       # full JSON trace

What you get: the mavis-ship skill + /mavis-ship slash command + mavis-orchestrate CLI on PATH. No MMAS, no agent-kernel.

Uninstall: ./install.sh --uninstall

Lane 3 — full local agent OS

npm install -g delegate-team                     # Lane 1 first
git clone https://github.com/imMamdouhaboammar/delegate-team
cd delegate-team
./install.sh --all --dry-run                      # preview what changes
./install.sh --all --trust-mode normal --yes      # do it
./install.sh --verify                             # confirm

Adds everything from Lane 2 plus:

• MMAS — multi-agent team runtime (Atlas + 7 specialists).
• agent-kernel — local memory, episodic recall, approval inbox, policy guard.
• Companion frameworks — superpowers, Waza, unslop-preflight, autoresearch.

Safety flags worth knowing:

| Flag | What it does | |---|---| | --dry-run | Show every write + network call without executing | | --no-network | Skip npx skills add, git clone, npm install -g | | --trust-mode strict | Disable MCP auto-load, block external downloads, print every sensitive op | | --trust-mode normal | Default — current safe behavior | | --yes | Non-interactive (for CI) |

Uninstall: ./install.sh --uninstall. Your ~/.agent-kernel/ memories are preserved.

🚀 Quick demo

# Lane 1: dispatch a task
dt run "Refactor the user model for multi-tenancy"

# Lane 2: route a task without executing
mavis-orchestrate "Build a pricing page with shadcn components"
# UI DELIVERY path — unslop audit is BLOCKING before /delegate-team.

# Lane 2: full JSON trace
dt route --explain "make API p95 < 200ms"
# {"task":"...","detected_signals":{...},"selected_workflow":"PERFORMANCE/METRIC",...}

# Lane 3: spawn a multi-agent team (with guardrails)
python3 ~/.mavis/agents/mavis/multi-agent/spawn-team.py \
    "Audit the auth layer" \
    --team atlas,scout,reviewer,sentinel \
    --timeout 900 --max-agents 4

🧩 Component map

| Component | What it does | Lane | Docs | |---|---|---|---| | dt CLI | Dispatch tasks to backends, run the proxy, check health | 1+ | DT.md | | /mavis-ship orchestrator | Natural-language task routing with structured traces | 2+ | docs/ROUTING.md | | agent-kernel | Local memory + governance (companion) | 3 | docs/AGENT-KERNEL-INTEGRATION.md | | MMAS | Multi-agent team runtime with guardrails | 3 | docs/MMAS.md | | Skill scaffolder | mavis-skill-scaffold CLI for new skills | 3 | scaffolder/ | | Backend agents | Codex, MiniMax, Gemini, MetaGPT — installed with Lane 3 | 3 | integrations/ |

For the conceptual architecture (what each layer is and what it is not), see docs/ARCHITECTURE.md.

🛡️ Safety

delegate-team runs AI agents on your local machine under your user account. It has built-in guards for the common risks:

• Workspace sandboxing (DT_WORKSPACE_ROOT)
• Command allowlist (override with DT_ALLOW_UNSAFE_COMMANDS=true)
• Supply chain guard (override with DT_ALLOW_DEP_INSTALL=true)
• MCP auto-load is off by default (DT_ENABLE_MCP=true to opt in)
• MMAS subprocess guardrails (max agents, timeouts, plan-only, kill switch)
• agent-kernel policy guard (pre-commit + Claude PreToolUse hooks)

Full threat model + opt-in switches: docs/SECURITY-MODEL.md.

📚 Documentation

| Doc | What's inside | |---|---| | docs/INSTALLATION.md | Lane-by-lane install + verify + uninstall | | docs/ARCHITECTURE.md | 5-layer mental model, data flow, boundaries | | docs/ROUTING.md | How the orchestrator picks a verdict | | docs/WORKFLOWS.md | Real examples per verdict | | docs/AGENT-KERNEL-INTEGRATION.md | Memory + governance boundary contract | | docs/MMAS.md | Multi-agent runtime + guardrails | | docs/SECURITY-MODEL.md | Threat model + opt-in switches | | INSTALL.md | Granular component-level install (Lane 2 / 3 detail) | | DT.md | Original dt CLI specifics | | CHANGELOG.md | v1 → v2.5 release notes | | SECURITY.md | Short vulnerability-reporting policy | | AGENTS.md | Repo conventions for contributors | | docs/audits/REPO-AUDIT-vNEXT.md | Pre-refactor audit (v2.6.0 baseline) |

📊 Project status

| Workflow | Status | Purpose | |---|---|---| | CI | | Build, typecheck, test, routing matrix, manifest validation | | npm publish | | Auto-publish on v* tag with OIDC provenance | | Release | | GitHub Release + tarball |

| Backend | Maturity | Notes | |---|---|---| | Codex (via god-agent) | Stable | opencode dispatcher | | MiniMax (via mmx) | Stable | minimax-coder/ | | Gemini (via google-genai) | Stable | vertex-coder/ | | MetaGPT team | Experimental | metagpt/ — multi-role team runner |

| Component | Maturity | Notes | |---|---|---| | dt CLI | Stable | npm package, ~21 KB tarball | | /mavis-ship orchestrator | Stable | bash + Skills.sh skill | | agent-kernel (vendored v0.0.5) | Stable | companion memory + governance | | MMAS multi-agent | Beta | runtime + guardrails (v2.6.0) |

🤝 Contributing

1. Read AGENTS.md for repo conventions.

2. Make a feature branch: git switch -c feat/your-change.

3. Run CI locally:

   bash -n install.sh
   bash -n orchestrator/scripts/orchestrate.sh
   python -m py_compile mmas/spawn-team.py mmas/watchdog.sh 2>/dev/null || python -m py_compile mmas/spawn-team.py mmas/hash-edit.py
   python -m json.tool < .claude-plugin/marketplace.json > /dev/null

4. Open a PR — CI runs the same checks on GitHub Actions runners.

Adding a new sub-skill? Use the scaffolder:

mavis-skill-scaffold --name my-new-skill --description "..." --type workflow

🛣️ Roadmap

• [x] v2.0.0 — supersystem release
• [x] v2.1.0 — Skills.sh + Claude Code marketplace compatibility
• [x] v2.1.1 — orchestrate.sh regex routing fixes
• [x] v2.2.0 — Polished README + expanded CI
• [x] v2.3.0 — npm package live + auto-publish
• [x] v2.4.0 — Polished README layout (distributed badges)
• [x] v2.5.0 — Bundled agent-kernel + memory stage in orchestrator
• [x] v2.5.1 — BundlePhobia Rspack fix in dt build path
• [x] v2.6.0 — Clarity + safety release: docs/ split, installer safety modes, routing traces, MMAS guardrails, agent-kernel boundary commands
• [ ] v2.6.1 — follow-ups from v2.6.0 (see CHANGELOG)
• [ ] v3.0.0 — multi-tenant team runs + per-project skill isolation

📜 License

MIT — see LICENSE.

💖 Acknowledgments

Built on top of:

• obra/superpowers — methodology layer
• tw93/Waza — 8-skills framework
• uditgoenka/autoresearch — metric loop
• imMamdouhaboammar/unslop-preflight — UI quality gate
• vercel-labs/skills — registry infrastructure
• imMamdouhaboammar/agent-kernel — vendored memory + governance

If this repo saved you time, star it ⭐ or open an issue.

Built by Mamdouh Aboammar · Cairo, Egypt 🇪🇬